An Analysis of the Possible Migration Routes of Oedaleus decorus asiaticus Bey-Bienko (Orthoptera: Acrididae) from Mongolia to China

Simple Summary Airflow is very important for the long-distance migration of O. decorus asiaticus, and wind shear, in particular, is the main factor related to forced landing. Analyzing the weather records, we found that the northwest wind prevailed when the population invaded. Specifically, from July to August, a large number of emerging adults appeared in the source areas of Mongolia, and the large-scale northwest wind was the decisive condition for the successful long-distance migration of O. decorus asiaticus. The species has a strong migratory ability, flying along the airflow for several nights. If the northwest air current meets the southwest warm current going north, a large number of O. decorus asiaticus will drop due to wind shear, and then a major outbreak will occur. Analysis of the source of the insects shows that the O. decorus asiaticus break outs in China may have originated from Mongolia. They were brought into China by the southerly airflow at night, and they likely made a forced landing in Beijing due to wind shear, sinking airflow, rainfall and other reasons. In summary, through analysis of the insect’s prevalence and the meteorological conditions in Mongolia, we can provide a basis for predicting the occurrence of O. decorus asiaticus in China. Abstract Oedaleus decorus asiaticus (Bey-Bienko) is a destructive pest in grasslands and adjacent farmland in northern China, Mongolia, and other countries in Asia. It has been supposed that this insect pest can migrate a long distance and then induce huge damages, however, the migration mechanism is still unrevealed. The current study uses insect light trap data from Yanqing (Beijing), together with regional meteorological data to determine how air flow contributes to the long-distance migration of O. decorus asiaticus. Our results indicate that sinking airflow is the main factor leading to the insects’ forced landing, and the prevailing northwest wind was associated with O. decorus asiaticus taking off in the northwest and moving southward with the airflow from July to September. Meanwhile, the insects have a strong migratory ability, flying along the airflow for several nights. Thus, when the airflow from the northwest met the northward-moving warm current from the southwest, a large number of insects were dropped due to sinking airflow, resulting in a large outbreak. Our simulations suggest that the source of the grasshoppers involved in these outbreaks during early 2000s in northern China probably is in Mongolia, and all evidence indicates that there are two important immigrant routes for O. decorus asiaticus migration from Mongolia to Beijing. These findings improves our understanding of the factors guiding O. decorus asiaticus migration, providing valuable information to reduce outbreaks in China that have origins from outside the country.


Introduction
Outbreaks of grasshoppers have had increasingly serious impacts on human activity in China over the past millennia. Oedaleus decorus asiaticus (O. decorus asiaticus) is an important pest affecting grasslands and adjacent farmland, and has caused huge damage to the livestock industry and ecological environment [1,2]. In China, O. decorus asiaticus mainly appears in Inner Mongolia, in typical grassland and desert steppe regions from the central and west of the Xilinguole League to the east of Ordos City [3]. During early July of 2002-2004, there were several sudden outbreaks of O. decorus asiaticus in northern China, including Chifeng City, Duolun County, Zhangjiakou City, Chengde City and Beijing [4]. We know that only adult insects have strong flight ability [5], with a maximum flight distance of 350 km in one night [6,7]. However, we found that the local population consists only of the fifth stage hoppers [4], and these cannot migrate over a long distance. Thus, we speculated that the source of O. decorus asiaticus involved in these outbreaks is likely to be an invasive population. However, the migration mechanism is still unrevealed.
Similar to factors affecting other migratory insects' flight ability and migration regularity [8], airflow has a strong influence on the success of O. decorus asiaticus' long-distance migration due to its inherently limited flight ability. Therefore, variations in seasonal airflow can have a critical effect on the annual migration of insects [9,10]. Weather factors, such as strong convergent airflows, can cause the forced landing of migratory individuals which terminates their flight, and can even lead to a large number of individuals landing in a concentrated manner, causing a large regional outbreak [11][12][13][14]. Thus, the causes of O. decorus asiaticus outbreaks and their migration patterns are closely related to climate factors. Characteristics of the weather can induce migration and have important impacts on flight. In particular, the insects can only take off on a large scale under suitable temperature and humidity conditions [15][16][17][18][19]. Swarming locusts usually move downwind so that the wind will take them to their destination [20]; suitable airflow takes them to a green vegetation habitat, providing them with food and new spawning opportunities.
Beijing is not a suitable habitat for the O. decorus asiaticus, and studies show that they cannot overwinter in Beijing [4]. However, there are reports of the occurrence of O. decorus asiaticus in this area every year, indicating that there is some immigration from other locations. In this study, we use Yanqing light trap data in combination with reanalyzed meteorological data available from the National Centers for Environmental Prediction (NCEP; United States of America) to develop a three-dimensional insect trajectory analysis. We used the Weather Research and Forecast model (WRF 3.9 model) [21] together with GrADS meteorological graphics software to simulate and visualize the migration paths of the O. decorus asiaticus migration into the Beijing area, and to identify the meteorological factors affecting these paths. Our study helps to clarify the migration regularity of the O. decorus asiaticus, and to determine the source and migration route of the small population of O. decorus asiaticus in Beijing. To develop our quantitative meteorological simulation, we used a terrain resolution of 2 , covering the world's Moderate Resolution Imaging Spectroradiometer (MODIS) and Terrain Gravity Wave Drag by Orography (GWDO) data. The NCEP FNL Global Tropospheric Analysis data was used to inform the initial field data and boundary conditions for the WRF model. After meteorological data are inputted, the model outputs an hourly forecast with 30 km × 30 km grid spacing which provides the conditions to allow us to calculate potential migration trajectories. The wind field data used in the trajectory calculation was derived from the WRF model, which is a new generation of mesoscale numerical weather prediction system. The model provides a high-resolution atmospheric background for the trajectory analysis of this study. The simulation area and parameter settings of the WRF model are shown in Table 1. Using the results of the WRF simulation and the pest situation records from agricultural stations, we selected the atmospheric field of the average nighttime high altitude flow field from the peak day of pest reduction, and selected the flow field map at 800 m altitude in the Beijing area. The weather background field of the O. decorus asiaticus flying at night was visualized using GrADS 2.1, and the influence of upper air flow on the migration and landing of O. decorus asiaticus population was analyzed.

Materials and Methods
To identify the potential sources of O. decorus asiaticus arriving in the Yanqing area, we simulated potential migratory routes for five nights backwards in time from each date that O. decorus asiaticus were captured in Yanqing. The trajectories were estimated with the following assumptions: (a) the migration speed at which the O. decorus asiaticus travels by air is the vector sum of the wind speed and its own flight speed, and its own flight speed is about 3 m/s [22][23][24]; (b) the migration time of the O. decorus asiaticus in the first night is from 20:00 p.m. on the night of take-off to their captured time, and the remaining four nights are from 20:00 p.m. to 3:00 a.m. of the next day, lasting for 7 h [4]; (c) O. decorus asiaticus migrates at an altitude of 500-1500 m above the ground [25]. Because the average altitude of the Mongolian Plateau is more than 1000 m, eight flight altitudes were used in this study: 750, 1000, 1250, 1500, 1750, 2000, 2250 and 2500 m above sea level [26]; (d) O. decorus asiaticus can migrate continuously for multiple nights. As mentioned above, the trajectory was estimated for the five consecutive nights prior to the night when the O. decorus asiaticus was captured, the landing points of the first night will serve as the departure points of the second night, and each departure point will have eight flying altitudes. In the next night, the end point of the trajectory of one night's travel was used as the starting point for the next day's take-off [24,27].
The trajectory simulation was carried out according to the high-altitude horizontal flow field and the above-mentioned biological parameters. In addition, we screened potential and effective migration trajectories based on characteristics, such as terrain, distribution area, and other biological characteristics [28]. The criteria for effective trajectories are as follows: (1) the time at the end of the trajectory should be in accordance with the takeoff rhythm of the O. decorus asiaticus; (2) there must be plants suitable for the feeding of O. decorus asiaticus at the end of the trajectory; (3) O. decorus asiaticus occurs in the trajectory termination area and can provide an effective source of emigration. Only when the flying altitude of O. decorus asiaticus is lower than the eight sea-level altitudes used (750, 1000, 1250, 1500, 1750, 2000, 2250 and 2500 m), can its flight trajectory can be maintained for five consecutive nights. Some tracks ended in less than five nights due to terrain factors ( Figure 1).

The Choice of Moving in Peak Day
By screening the trapping data from 2012 to 2017 and analyzing the start date of the immigration process of O. decorus asiaticus, we chose all the nights when the grasshoppers were captured as peak days, except for the ones that were captured between 8 p.m. to 9 p.m., because the grasshoppers taking off at 8 p.m. and landing at that time did not immigrate [19]. In total, we obtained 69 peak days (the date with trapped O. decorus asiaticus).

The Choice of Moving in Peak Day
By screening the trapping data from 2012 to 2017 and analyzing the start date of the immigration process of O. decorus asiaticus, we chose all the nights when the grasshoppers were captured as peak days, except for the ones that were captured between 8 p.m. to 9

Analysis of Population Dynamics under Light
The data of light traps from 2012 to 2017 showed that the period of light traps in Yanqing was from June and September and the peak period of light traps in Yanqing was from July and August, and longest trap time is 2017, lasting 31 days. In addition, while in recent years there were not many O. decorus asiaticus trapped by light traps in Yanqing from 2012 to 2017. The number was significantly higher in 2017 than in other years, with 566 trapped, with the largest number of O. decorus asiaticus captured from 8:00 p.m. on 15 August to 5:00 a.m. on 16 August 2017 (142). There was almost no light-captured grasshoppers in other months. O. decorus asiaticus hatch in May, and gradually emerge in July every year. Therefore, there is only successful light trapping of O. decorus asiaticus in Yanqing from July to August (Table 2).  The results of the trajectory analyses showed that while some trajectories terminated early (i.e., were less than five nights), most of the other trajectories come from the northwest, except for a small number of tracks from Southwest China. As can be seen in Figure 1 Of the calculated 14,297 trajectories, 12,819 (89.67%) were valid. There were 1478 trajectories that were terminated due to the high altitude of Inner Mongolia, accounting for 10.33% of the total trajectories (Table 3). Excluding the forcibly terminated trajectory landing points and invalid O. decorus asiaticus source landing points, there were a total of 12,833 effective trajectory landing points, which are mainly distributed in Inner Mongolia, Shanxi, and Hebei ( Figure 1).
These effective trajectories are mainly from Shanxi, Hebei, and Inner Mongolia, with a total of 10,382 (72.62%). Among them, the most effective trajectories were from Hebei with 5468, followed by Shanxi, Inner Mongolia, and Beijing, with 3112, 1802 and 1314, respectively. A small number of them came from Henan (414/2.90%), Liaoning (271/1.90%), Shaanxi (257/1.80%), Mongolia (49/0.34%), Tianjin (39/0.27%), Heilongjiang (39/0.27%), Gansu (20/0.14%), Jilin (10/0.07%), Shandong (9/0.06%), Chongqing (4/0.03%), Hubei (10/0.07%) and the sea surface (6/0.04%) (Tables 3 and 4).  Shanxi  139  21  193  9  336  38  142  3  146  11  2156  113  Hebei  444  78  42  52  598  90  583  157  79  20  3722  554  Henan  39 Table S1. Among them, landing on 24 days was associated with wind shear, 31 days were associated with sinking airflow, and 25 days were related to precipitation. It can be seen from the table that precipitation and sinking airflow are the key important factors for the landing of the O. decorus asiaticus in the Yanqing area. At this time, in the south, the cold air flow met a warm air flow greater than 6 m/s over Hebei and Beijing, resulting in wind shear ( Figure 2a) and a strong vertical airflow disturbance occurred on 7 July to 13 July. The speed of the sinking airflow was over 0.4 Pa/s (Figure 2b), which is sufficient to cause O. decorus asiaticus to land in the Yanqing area, with a resultant light trap peak in Yanqing at night. This sequence of events is supported by evidence from our backward trajectory simulation for 8 July 2017, which captured the effect of the Mongolian cyclone: the simulated trajectories came from the northwest, turned to the northeast, and the grasshopper could then enter Yanqing after passing through the central part of Inner Mongolia from outside Mongolia (Figure 1).   This sequence of events is supported by evidence from our backward trajectory simulation for 8 July 2017, which captured the effect of the Mongolian cyclone: the simulated trajectories came from the northwest, turned to the northeast, and the grasshopper could then enter Yanqing after passing through the central part of Inner Mongolia from outside Mongolia (Figure 1).

Dynamic Analysis of O. decorus asiaticus Migration in 2003
According

Discussion
Oedaleus decorus asiaticus is becoming an increasingly serious pest in the agricultural and pastoral ecotone in northern China. Due to its ability to cause a wide range of pastureland damage and engage in long distance migration, it is a difficult problem to monitor and control. Therefore, an understanding of their migration regularity is crucially im-

Discussion
Oedaleus decorus asiaticus is becoming an increasingly serious pest in the agricultural and pastoral ecotone in northern China. Due to its ability to cause a wide range of pastureland damage and engage in long distance migration, it is a difficult problem to monitor and control. Therefore, an understanding of their migration regularity is crucially important for the forecasting and control of many acridoid pests [29]. Trajectory regression analysis is one of the most common and effective methods to determine the possible source of migratory insects. It has been widely used in trajectory simulation analysis of migratory pests, such as Mythimna separata (Walker) (Noctuidae), Nilaparvata lugens (Stål) (Delphacidae), Cnaphalocrocis medialis (Guenee) (Crambidae), and Spodoptera frugiperda (Smith) (Noctuidae) [30][31][32][33]. Because insects are small and have limited flight capabilities, their migration ability cannot be compared with that of mammals, birds and other large animals. However, studies on the migration strategies of Noctuidae adults, such as Helicoverpa armigera (Hubner) and Argyrogramma agnata (Staudinger), have shown that insects with the ability to migrate autonomously can drift farther with the wind than inert particles in the air. This reflects insects' use of an adaptive migration strategy related to the wind temperature field of the atmospheric boundary layer [9,21]. That is, to achieve long-distance migration, insects often use suitable wind fields to help them migrate [34].
There was no initial record of O. decorus asiaticus distribution in Beijing. In 2002, Jiang Xiang et al. kept the collected O. decorus asiaticus in the laboratory and found that while some of them could survive for a short time, the long-term survival rate of adults was extremely low, and the number of eggs laid was very small [4]. This finding indicated that O. decorus asiaticus were not coming from a local source, suggesting the likelihood of long-distance migration. That is, it seemed likely that most of the O. decorus asiaticus caught in Beijing migrated from outside populations.
In order to explore its migration process, we assumed a flight capability of 3.0 m/s [25] when simulating the migration trajectory of the O. decorus asiaticus, and used the eight flight altitude parameters of 750, 1000, 1250, 1500, 1750, 2000, 2250, and 2500 m. Through trajectory analysis, we concluded that the flight altitude is more than 1250 m. This is because the average altitude of Inner Mongolia is above 1000 m, so lower altitude trajectories will terminate in fewer than five nights, before the insects could have reached Beijing. The parameters of take-off and landing time, continuous flight time, migration times and flight altitude were set to accurately simulate the origin of O. decorus asiaticus in the Beijing area. According to the trajectory of light-trapped adults from 2012-2017, it is clear that most of the O. decorus asiaticus sources are in Hebei, Shanxi, Inner Mongolia, and Beijing, with a smaller proportion originating in Henan, Liaoning, and Shaanxi. The eastern part of Mongolia can also be the source of O. decorus asiaticus in China. Since there is no O. asiaticus distribution record in Chongqing, Hubei and the sea surface, the landing point is invalid [3]. The migration of O. decorus asiaticus does not involve completely passively drifting with the wind, but to a certain extent the insects autonomously choose to travel with the wind, migrating when conditions are suitable [19]. Insects migrating with the wind will intensively land under meteorological conditions, such as heavy rainfall, sinking air currents, and combined wind directions. They will also actively land when their energy materials are exhausted or the temperature drops below their flight threshold [34][35][36]. Among these factors, precipitation and sinking airflow are two important meteorological factors that affect the landing of aerial insect clusters [13]. For example, Guanheng Jiang et al. analyzed the meteorological factors during the 67 northward migrations and 15 southward migrations of the brown planthopper from 1977 to 1978, and found that precipitation and sinking airflow were the main meteorological factors affecting the large-scale landing of the brown planthopper [37]. Bao Yunxuan et al. analyzed the migration process of Sogatella furcifera (Horváth) (Delphacidae) that appeared on 10-11 July 2003 and found that rainfall was the direct cause of the concentrated landing of this planthopper species [38]. In the current study, we analyzed likely landing mechanisms on the 69 peak days of O. ecorus asiaticus migration into Yanqing area in July to September from 2012-2017 (Table S1); 24 times were related to wind shear and 31 times were related to the down draft, and 25 times were related to precipitation. Previous studies have shown that wind shear can cause large-scale landings of migratory insects [36][37][38]. In this study, 24 landings of O. decorus asiaticus were related to wind shear. Wind shear causes changes in the wind speed and direction of the airflow, and the sudden change in temperature can interrupt the southward migration of the O. decorus asiaticus, instigating a concentrated landing. Cold air from Siberia flows southward, and meets the warm air coming from the south, forming a long and narrow convergence zone. The wind shear thus formed becomes the main factor that causes O. decorus asiaticus to land. The Yanqing area belongs to a temperate monsoon climate, which is characterized by hot and rainy summers and cold and dry winters. Our results showed that the 25 selected immigration peaks all appeared in July or August, synchronized with the rainy season. Rainfall is an important factor influencing the concentration and decline of populations of O. decorus asiaticus in Yanqing. Because the terrain of Beijing and Hebei is high in the northwest and low in the southeast, gradually descending from northwest to southeast, the vertical disturbance caused by the decline of the terrain during the movement of the airflow also influences O. decorus asiaticus landing patterns. In addition, there are 17 landing times of O. decorus asiaticus which were not found to be related to rainfall, wind shear and downdraft, which may have occurred because their energy materials were exhausted, but the specific reasons need further investigation.
The genus Oedaleus originated in Ethiopia, Africa. Most of its species are closely related to locusts [39][40][41]. Based on mitochondrial DNA analysis, the genus Oedaleus may be the ancestor of locusts [41,42]. Although O. decorus asiaticus is a grasshopper, it also has the ability of locust's long-distance migration. Therefore, we can refer to locust migration to study the migration of O. decorus asiaticus. Moreover, previous studies found that O. decorus asiaticus has migration behavior, and temperature and wind can significantly affect the flight of O. decorus asiaticus at night [19]. O. decorus asiaticus can fly 150 km per night with the help of wind [19]. Beijing is about 600 km away from Mongolia, so O. decorus asiaticus in Mongolian can reach Beijing in less than five nights. Through light trapping and trajectory simulation analysis, it is found that O. decorus asiaticus mostly migrate downwind at night, which is consistent with the flight behavior of the Senegalese Oedaleus senegalensis [43].
On 12 July 2002, it was reported that a large number of O. decorus asiaticus appeared in Beijing. Jiang Xiang et al. conducted a telephone survey and found that a large number of O. asiaticus appeared in Chengde City in Hebei Province, Zhangjiakou City in Hebei Province, and Beijing between 10 and 12 July 2002 [4]. Because most of the O. asiaticus distributed in China in early July were elderly nymphs [44], the source of the O. asiaticus is likely to be from abroad. Therefore, the wind direction at night at that time shows that grasshoppers followed the airflow from Mongolia into China and landed in Hebei province (  (Figures 3 and 4), we can summarize the possible migration routes of O. decorus asiaticus after entering China from Mongolia as follows: (1) entering Erlianhaote, Xianghuang Banner, Zhangjiakou City, and Hebei Province through the west Sunite Banner of Inner Mongolia, and continuing to Yanqing City, Beijing; (2) entering China via the Abaga Banner Xilingol league, Xilinhot in Inner Mongolia and continuing to move south to Yanqing. At the same time, we mainly analyzed the influence of the Mongolian cyclone center on the migration of O. decorus asiaticus. There are also extratropical cyclone centers in adjacent areas, such as Eastern Inner Mongolia and Heilongjiang [45,46]. However, combined with field observation data, there were no adult O. decorus asiaticus in these areas in the beginning and middle of July, which needs to be further studied.

Conclusions
Oedaleus decorus asiaticus is known from Russia (southern Siberia including the Tyva Republic, the Republic of Buryatia, and Transbaikal region), the Mongolian People's Republic, and the People's Republic of China, but sometimes is considered as a subspecies of widely distributed O. decoratus [40,47]. O. decorus asiaticus is an important pest affecting the farming pastoral ecotone of northern China, which has caused great losses to agriculture and animal husbandry [3,4,44]. The results of the study showed that airflow is very important for the long-distance migration of O. decorus asiaticus, and that wind shear, in particular, was the main factor related to its forced landing. Analyzing the weather records, we found that the northwest wind prevailed when the population invaded. Specifically, from July to August, a large number of emerging adults appeared in the source areas of Mongolia, and the large-scale northwest wind was the decisive condition for the successful long-distance migration of O. decorus asiaticus. It has a strong migratory ability, flying along the airflow for several nights. If the northwest air current meets the southwest warm current going north, a large number of O. decorus asiaticus will drop due to wind shear, and then a major outbreak will occur. The analysis of the source of the insects shows that the O. decorus asiaticus break outs in China may have originated from Mongolia. They were brought into China by the southerly airflow at night, and they could make a forced landing in Beijing due to wind shear, sinking airflow, rainfall and other reasons. In sum, through the analysis of the insect prevalence and meteorological conditions in Mongolia, we can provide a basis for predicting the occurrence of O. decorus asiaticus in China, and the possible migration routes of O. decorus asiaticus after entering China from Mongolia are as follows: (1) [6,48]. Therefore, In the main invasion season in summer, we should make full use of high-altitude searchlights, insect radar and other pest monitoring technologies to improve monitoring and early warning of O. decorus asiaticus appearances in northern China.